## Estimating the rate of return on Social Security

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Topic Author
DonIce
Posts: 740
Joined: Thu Feb 21, 2019 6:44 pm

### Estimating the rate of return on Social Security

I was thinking about this and decided to run some numbers. The average individual income is around \$50k (round numbers). Lets say one works from 25 to 65 at that wage, and their salary always goes up matching inflation, so we can call it 40 years of working at a salary of \$50k. Everything below can be assumed to be in real (inflation adjusted) dollars. And let's assume social security keeps going up with inflation. And lets assume you then live to the average life expectancy in the US, which is 79. So you collect for 14 years.

The total amount you pay in is 40 years * \$50,000 * 6.2% = \$124,000. Your employer contributes another \$124,000, for a total of \$248,000.

Running a simple calculator for expected social security payout, it says if you've had a \$50,000 income, you'll get a payout of \$15,176 per year. Collecting this payout for 14 years, you receive a total of \$212,464.

Given that you've been contributing \$6200 per year for 40 years, that means your real return over that period (above inflation) has been -0.8% (yes, negative). If you assume longer life expectancy you can eventually get a positive return, but the rate of return is still very low. Did I make a mistake somewhere? Cause that seems like an unreasonably bad return.

dkturner
Posts: 1526
Joined: Sun Feb 25, 2007 7:58 pm

### Re: Estimating the rate of return on Social Security

Since Social Security benefits are regressive (the higher your income, the lower your benefit in percentage of income terms) the calculation is different for each participant. A participant with an average income of \$20,000 receives a much higher percentage of income benefit than a participant with an \$80,000 average income. Low income participants are well rewarded - at the expense of high income participants. Someone has to pay the piper.

IowaFarmBoy
Posts: 729
Joined: Fri Jan 22, 2010 8:19 am

### Re: Estimating the rate of return on Social Security

I've been struggling with whether you can assume away inflation and the COLA benefits the way you have but I've almost convinced myself that you could do that if there is no inflation. However, that is far from the reality of the past. It would be interesting to run this analysis including inflation in earnings and benefits. Kitces has done some analysis (which I haven't read carefully) and comes up with a real return of .9%- roughly equal to TIPS https://www.kitces.com/blog/social-secu ... t-roi-irr/

There are several other things that come to mind:
• The average life expectancy for a 65 year old is around 19-20 years more. I suspect your 14 years more number includes everyone that died before that point. This would mean 5-6 years more payout.
• If you have a non-working spouse, your total benefit will be 50% higher. And the likelihood of at least one of you living longer is higher.
• There are other SS benefits beyond the retirement check- disability coverage, survivors' benefits, etc. which have a cost and provide value

KlangFool
Posts: 14696
Joined: Sat Oct 11, 2008 12:35 pm

### Re: Estimating the rate of return on Social Security

DonIce wrote:
Fri May 03, 2019 12:57 pm
I was thinking about this and decided to run some numbers. The average individual income is around \$50k (round numbers). Lets say one works from 25 to 65 at that wage, and their salary always goes up matching inflation, so we can call it 40 years of working at a salary of \$50k. Everything below can be assumed to be in real (inflation adjusted) dollars. And let's assume social security keeps going up with inflation. And lets assume you then live to the average life expectancy in the US, which is 79. So you collect for 14 years.

The total amount you pay in is 40 years * \$50,000 * 6.2% = \$124,000. Your employer contributes another \$124,000, for a total of \$248,000.

Running a simple calculator for expected social security payout, it says if you've had a \$50,000 income, you'll get a payout of \$15,176 per year. Collecting this payout for 14 years, you receive a total of \$212,464.

Given that you've been contributing \$6200 per year for 40 years, that means your real return over that period (above inflation) has been -0.8% (yes, negative). If you assume longer life expectancy you can eventually get a positive return, but the rate of return is still very low. Did I make a mistake somewhere? Cause that seems like an unreasonably bad return.
DonIce,

If you collect at 65 years old, the number is \$2,814 per month or \$33,773 per year.

KlangFool

KlangFool
Posts: 14696
Joined: Sat Oct 11, 2008 12:35 pm

### Re: Estimating the rate of return on Social Security

IowaFarmBoy wrote:
Fri May 03, 2019 1:23 pm
I've been struggling with whether you can assume away inflation and the COLA benefits the way you have but I've almost convinced myself that you could do that if there is no inflation. However, that is far from the reality of the past. It would be interesting to run this analysis including inflation in earnings and benefits. Kitces has done some analysis (which I haven't read carefully) and comes up with a real return of .9%- roughly equal to TIPS https://www.kitces.com/blog/social-secu ... t-roi-irr/

There are several other things that come to mind:
• The average life expectancy for a 65 year old is around 19-20 years more. I suspect your 14 years more number includes everyone that died before that point. This would mean 5-6 years more payout.
• If you have a non-working spouse, your total benefit will be 50% higher. And the likelihood of at least one of you living longer is higher.
• There are other SS benefits beyond the retirement check- disability coverage, survivors' benefits, etc. which have a cost and provide value
IowaFarmBoy,

<<Kitces has done some analysis (which I haven't read carefully) and comes up with a real return of .9%- roughly equal to TIPS >>

We need to discuss the bend points in order for this to be useful. Before the first bend point, it is a very good deal. Between the first and second bend point not so much. After the second bend point, probably the return is not there.

For the taxed 100K before first bend point, you are paid \$214.50 per month for \$7,650 tax that you paid.

For the taxed 100K After first bend point and before the second bend point, you are paid \$76.19 per month for \$7,650 tax that you paid.

For the taxed 100K After the second bend point, you are paid \$35.71 per month for \$7,650 tax that you paid.

KlangFool

Topic Author
DonIce
Posts: 740
Joined: Thu Feb 21, 2019 6:44 pm

### Re: Estimating the rate of return on Social Security

KlangFool wrote:
Fri May 03, 2019 1:23 pm
DonIce,

If you collect at 65 years old, the number is \$2,814 per month or \$33,773 per year.

KlangFool
My link is just to the calculator, not to any particular number (it doesn't generate custom links with specific data entered into the fields, as far as I can tell). I just re-checked. If you enter current age and retirement age as 65 and annual income as \$50,000, it gives you \$15176 per year. If you enter current age as 25 then you get different numbers for the future depending on what you assume the inflation rate to be, but you can simplify all that by just assuming that you are retiring today in today's dollars.

KlangFool
Posts: 14696
Joined: Sat Oct 11, 2008 12:35 pm

### Re: Estimating the rate of return on Social Security

DonIce wrote:
Fri May 03, 2019 1:35 pm
KlangFool wrote:
Fri May 03, 2019 1:23 pm
DonIce,

If you collect at 65 years old, the number is \$2,814 per month or \$33,773 per year.

KlangFool
My link is just to the calculator, not to any particular number (it doesn't generate custom links with specific data entered into the fields, as far as I can tell). I just re-checked. If you enter current age and retirement age as 65 and annual income as \$50,000, it gives you \$15176 per year. If you enter current age as 25 then you get different numbers for the future depending on what you assume the inflation rate to be, but you can simplify all that by just assuming that you are retiring today in today's dollars.
DonIce,

I can tell you that your number is wrong. I have my own spreadsheet to calculate this kind of stuff. It is at least 2K per month.

It is very simple to check. Take 50K X 40 = 2 million. Take the 2 million divided by 35 years and 12 months. You get the PIA. Use the PIA and bend points to calculate the benefit numbers, you get 2K per month.

https://www.ssa.gov/oact/cola/bendpoints.html

This simple check will generate an estimate lowered than the actual number.

KlangFool

sandramjet
Posts: 283
Joined: Thu Oct 23, 2014 11:28 pm

### Re: Estimating the rate of return on Social Security

IMHO, the problem with all of this is that it fundamentally is thinking about SS as an investment, while you should view it as insurance.

Do you calculate the return on your homeowners or car insurance? ( If I did, it would be way more "negative" than SS. ) Instead, ask yourself what would it cost you to get some disability insurance, and longevity insurance, all with inflation adjustments?

Topic Author
DonIce
Posts: 740
Joined: Thu Feb 21, 2019 6:44 pm

### Re: Estimating the rate of return on Social Security

KlangFool wrote:
Fri May 03, 2019 1:41 pm
DonIce,

I can tell you that your number is wrong. I have my own spreadsheet to calculate this kind of stuff. It is at least 2K per month.

It is very simple to check. Take 50K X 40 = 2 million. Take the 2 million divided by 35 years and 12 months. You get the PIA. Use the PIA and bend points to calculate the benefit numbers, you get 2K per month.

https://www.ssa.gov/oact/cola/bendpoints.html

This simple check will generate an estimate lowered than the actual number.

KlangFool
With a \$50k income, your average monthly income is \$4167.

The first bendpoint is \$926 and the 2nd bendpoint is \$5583.

The monthly benefit is 90% of 926 + 32% of (4167-926) = \$833.4 + \$1037.12 = \$1870.52. But that's at full retirement age which is 67 whereas I used 65 above. Still, you're right, seems like the number I was getting from the calculator was wrong.

By the way, I think the sentence of yours I bolded above is wrong? You can't take 40 years of income and divide them into 35 years? Doesn't the formula just take your highest 35 years and throw out the rest?

Dottie57
Posts: 7556
Joined: Thu May 19, 2016 5:43 pm
Location: Earth Northern Hemisphere

### Re: Estimating the rate of return on Social Security

sandramjet wrote:
Fri May 03, 2019 1:42 pm
IMHO, the problem with all of this is that it fundamentally is thinking about SS as an investment, while you should view it as insurance.

Do you calculate the return on your homeowners or car insurance? ( If I did, it would be way more "negative" than SS. ) Instead, ask yourself what would it cost you to get some disability insurance, and longevity insurance, all with inflation adjustments?
+1

KlangFool
Posts: 14696
Joined: Sat Oct 11, 2008 12:35 pm

### Re: Estimating the rate of return on Social Security

DonIce wrote:
Fri May 03, 2019 1:52 pm
KlangFool wrote:
Fri May 03, 2019 1:41 pm
DonIce,

I can tell you that your number is wrong. I have my own spreadsheet to calculate this kind of stuff. It is at least 2K per month.

It is very simple to check. Take 50K X 40 = 2 million. Take the 2 million divided by 35 years and 12 months. You get the PIA. Use the PIA and bend points to calculate the benefit numbers, you get 2K per month.

https://www.ssa.gov/oact/cola/bendpoints.html

This simple check will generate an estimate lowered than the actual number.

KlangFool
With a \$50k income, your average monthly income is \$4167.

The first bendpoint is \$926 and the 2nd bendpoint is \$5583.

The monthly benefit is 90% of 926 + 32% of (4167-926) = \$833.4 + \$1037.12 = \$1870.52. But that's at full retirement age which is 67 whereas I used 65 above. Still, you're right, seems like the number I was getting from the calculator was wrong.

By the way, I think the sentence of yours I bolded above is wrong? You can't take 40 years of income and divide them into 35 years? Doesn't the formula just take your highest 35 years and throw out the rest?
DonIce,

Yes, I made a mistake. But, that mistake probably will be offset by the wage adjustment upward. And, the number will come up to be about 2K per month.

KlangFool

KlangFool
Posts: 14696
Joined: Sat Oct 11, 2008 12:35 pm

### Re: Estimating the rate of return on Social Security

sandramjet wrote:
Fri May 03, 2019 1:42 pm
IMHO, the problem with all of this is that it fundamentally is thinking about SS as an investment, while you should view it as insurance.

Do you calculate the return on your homeowners or car insurance? ( If I did, it would be way more "negative" than SS. ) Instead, ask yourself what would it cost you to get some disability insurance, and longevity insurance, all with inflation adjustments?
sandramjet,

In some context, it might be useful. I plan to early retire. So, I could calculate the impact of my early retirement and/or working one more year on my social security benefit. In summary, after the second bend point, the impact on social security benefit is minimal.

KlangFool

JoeRetire
Posts: 4523
Joined: Tue Jan 16, 2018 2:44 pm

### Re: Estimating the rate of return on Social Security

DonIce wrote:
Fri May 03, 2019 12:57 pm
Lets say one works from 25 to 65 at that wage, and their salary always goes up matching inflation, so we can call it 40 years of working at a salary of \$50k.
The social security impact of 50k many years ago is much higher than 50k this year.

The simplistic tool you linked to doesn't take that into account.

Using this tool, you are underestimating your social security benefits, likely by a large amount. And thus your "rate of return" is wrong.

And of course you are assigning \$0 in value to all the other aspects of social security (spousal, survivor, etc, etc.)
Very Stable Genius

GrowthSeeker
Posts: 808
Joined: Tue May 15, 2018 10:14 pm

### Re: Estimating the rate of return on Social Security

Here is the methodology I would use if I wanted to know "how good an investment" SS has been for me.
I would enter from my SS report, the annual "SS" income for each year of my history.
I would then attempt to reconstruct the amount of money I paid into SS each year.
Once source I found for this was here .
One problem is that prior to 1991, there was a single payment to FICA (SS and Medicare were combined), so I'm not sure how much was attributed to SS.
Another issue would be whether to include the amount the employer paid, so I would also reconstruct that data and then solve the problem both ways, both including and excluding employer payments.
I would set up a spreadsheet where the starting balance is zero, and each year the balance increases by the amount I paid into SS times (1 + ROI) where ROI is a guess at the return on investment for my mythical SS account.

Since I am already receiving SS, I know the amount of SS I receive.
I would take the year I first received SS and try to compute the cost of a SPIA indexed for inflation for my same age and sex. That becomes, for me, the final value of my theoretical SS "account".

Then use the spreadsheet's "Goal Seek" function to make the final value of the account match up to the amount needed to purchase the SPIA by varying the ROI.

That's how I would do it. The answer is going to be different for each individual. Looking at it like a SPIA avoids the need to calculate different numbers for different life expectancies.
Just because you're paranoid doesn't mean they're NOT out to get you.

rkhusky
Posts: 7891
Joined: Thu Aug 18, 2011 8:09 pm

### Re: Estimating the rate of return on Social Security

JoeRetire wrote:
Fri May 03, 2019 2:10 pm
DonIce wrote:
Fri May 03, 2019 12:57 pm
Lets say one works from 25 to 65 at that wage, and their salary always goes up matching inflation, so we can call it 40 years of working at a salary of \$50k.
The social security impact of 50k many years ago is much higher than 50k this year.
I think the assumption is that the individual made a yearly wage such that if one applies the SS inflation adjustments (https://www.ssa.gov/oact/cola/awifactors.html), then one ends up with \$50K for every year.

#Cruncher
Posts: 2869
Joined: Fri May 14, 2010 2:33 am
Location: New York City
Contact:

### Re: Estimating the rate of return on Social Security

I ran an analysis using Case A of the SSA's Benefit Calculation Examples for Workers Retiring in 2019, adjusted for CPI-U, and applying double the OASDI Tax Rates. The Primary Insurance Amount (PIA) payable starting at age 66.5 is \$1,879.80 (see Case A page 2) or \$22,588 per year (in 2019 dollars). I get the following Internal Rates of Return assuming one lives to age:

Code: Select all

``````75  (0.1%)
80   1.3%
85   2.1%
90   2.6%``````
I adjusted all the salaries to 2019 dollars so these are "real" returns, like those on TIPS. Here is the complete table. Internal Rate of Return is calculated with the Excel IRR function.

Code: Select all

``````             Cash                             Adj     Tax
Year  Age    Flow   IRR    Salary   CPI-U   Salary   Rate``````

Code: Select all

``````1979   22  (3,971)         10,733   69.800  39,088  10.16%
1980   23  (3,784)         11,737   80.100  37,248  10.16%
1981   24  (3,983)         12,959   88.500  37,223  10.70%
1982   25  (3,984)         13,715   94.500  36,893  10.80%
1983   26  (4,046)         14,428   97.900  37,463  10.80%
1984   27  (4,328)         15,324  102.600  37,967  11.40%
1985   28  (4,365)         16,027  106.400  38,290  11.40%
1986   29  (4,409)         16,553  108.800  38,675  11.40%
1987   30  (4,566)         17,663  112.100  40,053  11.40%
1988   31  (4,916)         18,590  116.500  40,563  12.12%
1989   32  (4,884)         19,386  122.300  40,294  12.12%
1990   33  (4,982)         20,343  128.700  40,181  12.40%
1991   34  (4,942)         21,166  135.000  39,855  12.40%
1992   35  (5,052)         22,324  139.300  40,738  12.40%
1993   36  (4,957)         22,584  143.600  39,978  12.40%
1994   37  (4,981)         23,260  147.200  40,168  12.40%
1995   38  (5,052)         24,265  151.400  40,741  12.40%
1996   39  (5,168)         25,529  155.700  41,680  12.40%
1997   40  (5,339)         27,099  160.000  43,054  12.40%
1998   41  (5,558)         28,602  162.200  44,825  12.40%
1999   42  (5,786)         30,286  165.000  46,659  12.40%
2000   43  (5,906)         32,056  171.100  47,625  12.40%
2001   44  (5,889)         32,918  176.200  47,490  12.40%
2002   45  (5,879)         33,346  178.800  47,408  12.40%
2003   46  (5,863)         34,261  184.200  47,281  12.40%
2004   47  (6,048)         35,959  187.400  48,777  12.40%
2005   48  (6,096)         37,384  193.300  49,162  12.40%
2006   49  (6,187)         39,216  199.800  49,894  12.40%
2007   50  (6,311)         41,115  205.352  50,896  12.40%
2008   51  (6,227)         42,183  213.528  50,218  12.40%
2009   52  (6,175)         41,667  212.709  49,795  12.40%
2010   53  (6,195)         42,774  217.631  49,962  12.40%
2011   54  (6,240)         44,241  223.467  50,326  12.40%
2012   55  (6,287)         45,754  229.392  50,703  12.40%
2013   56  (6,293)         46,471  232.773  50,749  12.40%
2014   57  (6,437)         48,257  236.293  51,914  12.40%
2015   58  (6,685)         50,078  236.119  53,913  12.40%
2016   59  (6,723)         50,787  238.132  54,214  12.40%
2017   60  (6,812)         52,690  243.801  54,938  12.40%
2018   61  (6,882)         54,489  249.554  55,504  12.40%
2019   62                          254.202          0.00%
2020   63
2021   64
2022   65
2023   66  11,279
2024   67  22,558
2025   68  22,558
2026   69  22,558
2027   70  22,558
2028   71  22,558
2029   72  22,558
2030   73  22,558
2031   74  22,558  (0.5%)
2032   75  22,558  (0.1%)  <--
2033   76  22,558   0.3%
2034   77  22,558   0.6%
2035   78  22,558   0.9%
2036   79  22,558   1.1%
2037   80  22,558   1.3%   <--
2038   81  22,558   1.5%
2039   82  22,558   1.7%
2040   83  22,558   1.9%
2041   84  22,558   2.0%
2042   85  22,558   2.1%   <--
2043   86  22,558   2.2%
2044   87  22,558   2.3%
2045   88  22,558   2.4%
2046   89  22,558   2.5%
2047   90  22,558   2.6%   <--
2048   91  22,558   2.7%
2049   92  22,558   2.8%
2050   93  22,558   2.8%
2051   94  22,558   2.9%
2052   95  22,558   3.0%
2053   96  22,558   3.0%
2054   97  22,558   3.1%
2055   98  22,558   3.1%
2056   99  22,558   3.2%
2057  100  22,558   3.2%``````

jeffyscott
Posts: 8442
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

### Re: Estimating the rate of return on Social Security

A couple of issues, life expectancy at 65 is about 20 years, not 14 years. This means average pay out would be ~\$304,000 over 20 years.

https://www.ssa.gov/planners/lifeexpectancy.html

The 6.2% funds OASDI, which stands for old age, survivors, and disability insurance. Of the 6.2%, 0.9% funds DI so that needs to be completely excluded when doing your evaluation. Of the remaining 5.3%, they do not specify how much is for "old age" and how much is for survivors benefits. So that means something less than \$212,000 paid in toward retirement benefits.

https://www.ssa.gov/oact/progdata/oasdiRates.html
Time is your friend; impulse is your enemy. - John C. Bogle

Topic Author
DonIce
Posts: 740
Joined: Thu Feb 21, 2019 6:44 pm

### Re: Estimating the rate of return on Social Security

jeffyscott wrote:
Fri May 03, 2019 5:06 pm
A couple of issues, life expectancy at 65 is about 20 years, not 14 years. This means average pay out would be ~\$304,000 over 20 years.

https://www.ssa.gov/planners/lifeexpectancy.html

The 6.2% funds OASDI, which stands for old age, survivors, and disability insurance. Of the 6.2%, 0.9% funds DI so that needs to be completely excluded when doing your evaluation. Of the remaining 5.3%, they do not specify how much is for "old age" and how much is for survivors benefits. So that means something less than \$212,000 paid in toward retirement benefits.

https://www.ssa.gov/oact/progdata/oasdiRates.html
Thanks, good points on excluding the disability and survivor benefits from the calculation. Regarding life expectancy, I thought it made more sense to use overall life expectancy rather than life expectancy at age 65. From the perspective of someone starting their career and contemplating the rate of return on social security contributions, they haven't reached age 65 yet.

Topic Author
DonIce
Posts: 740
Joined: Thu Feb 21, 2019 6:44 pm

### Re: Estimating the rate of return on Social Security

#Cruncher wrote:
Fri May 03, 2019 5:03 pm
I ran an analysis using Case A of the SSA's Benefit Calculation Examples for Workers Retiring in 2019, adjusted for CPI-U, and applying double the OASDI Tax Rates. The Primary Insurance Amount (PIA) payable starting at age 66.5 is \$1,879.80 (see Case A page 2) or \$22,588 per year (in 2019 dollars). I get the following Internal Rates of Return assuming one lives to age:
Thanks for crunching the numbers! Interesting to see. Why does your data have a gap though where earnings stop at age 62 but social security income doesn't begin until age 66.5 though? I understand that some people may choose to retire at age 62, draw on other resources, and then start taking SS at age 66.5, but that doesn't strike me as what the baseline assumption should necessarily be?

willthrill81
Posts: 15190
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

### Re: Estimating the rate of return on Social Security

I think it should be pointed out that the average Boglehead has a greater life expectancy than the general population since higher levels of wealth are associated with longer life expectancies. But how much longer they are is difficult to nail down. Suffice it to say that the SSA life expectancies are probably on the low side of what's reasonable, ceterus paribus. This also means that the rate of return for a typical Boglehead is probably a little higher than some of the estimates used so far in this thread.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

Wakefield1
Posts: 1059
Joined: Mon Nov 14, 2016 10:10 pm

### Re: Estimating the rate of return on Social Security

40 years ago, \$50,000 per year was very good money to be making.

Dontridetheindexdown
Posts: 92
Joined: Sat Feb 14, 2015 10:08 pm

### Re: Estimating the rate of return on Social Security

I receive Social Security OASDI (Old Age, Survivors and Disability Insurance) as an old age annuitant. My payments to OASDI began when I was still in high school, my payout began 45 years later.

I ran the numbers to determine the "pivot point." It is 5.38%. If every dollar I had paid in was invested at 5.38% per annum, I will receive 5.38% of that total amount each year, adjusted upward for inflation but never adjusted downward for deflation, until I die.

At that time, the total investment is "forfeited" back into the risk pool, to provide benefits to survivors of wage earners, and to the disabled. This is a wonderful system!

Many people are afflicted by illness or lost-time injury during their working years. I was extremely fortunate to have been spared.

If the investment had increased at a higher rate than 5.38%, I would be receiving a lower percentage of the total. Similarly, if the investment had grown more slowly, I would be receiving a higher percentage of the total.

During those 45 years, the measured inflation rate averaged 4.41%. My "payout" is .96% greater than measured inflation!

(Rounding errors obtain. My payout is actually 5.376%, inflation was 4.413%, payout is .963% higher than inflation. The miracle is that with this annuity, I stored away a portion of my productivity. Because the return on investment exceeded inflation, it is worth more now, when I am no longer competitively productive, than it was worth when I stored it away.)

Even more wonderful, if I predecease my wife, she will receive the greater of my OASDI survivor benefit, or her own old age benefit. Similarly, if she predeceases me, I will receive the greater of her OASDI survivor benefit, or my old age benefit.

acegolfer
Posts: 1570
Joined: Tue Aug 25, 2009 9:40 am

### Re: Estimating the rate of return on Social Security

How did you get -0.8%?

Using your numbers + =IRR(), I got -0.578% (with \$6200 contribution) or 1.934% (w/ \$3100 contribution) annualized return.

GrowthSeeker
Posts: 808
Joined: Tue May 15, 2018 10:14 pm

### Re: Estimating the rate of return on Social Security

I used the method I outlined above, about 10 posts up.
* I used my own personal Social Security report to find the "SS Income" for each year that I worked.
* For historical SS tax rates, I used this website. (SS and Medicare are listed separately there).
* For inflation, I used the same website used by #cruncher above. Specifically, I used the annual rate-of-change (March to March) in CPI-U which was how they calculated their "1 year" column.
* I decided it was more fair to include both the employee AND employer "contributions".
* I used immediateannuities.com to find that the payout rate for a 66 yo male was 6.83%, but if indexed for CPI, the payout rate was 4.80%

I let my fictitious SS account grow in two ways, first at a constant rate and secondly as a constant amount above inflation. In each case, my SS account would have to grow to equal "my monthly SS benefit" * 12 / 4.8% as of the year I turned 66.

Results:
For my annual payments into SS (both employee and employer) to have grown to an amount identical to the purchase price of a CPI-adjusted SPIA giving me the same SS monthly payments for life, the ROI would have to be either:
A constant (nominal) 4.76%.
or
A (real) 2.37%, ie greater than inflation each year.

----
[Edit: I wonder what my SS benefit would be if I could have put the same dollar amount into a BH-style portfolio each year. That's an exercise for another day.]
Just because you're paranoid doesn't mean they're NOT out to get you.

#Cruncher
Posts: 2869
Joined: Fri May 14, 2010 2:33 am
Location: New York City
Contact:

### Re: Estimating the rate of return on Social Security

DonIce wrote:
Fri May 03, 2019 5:23 pm
Why does your data have a gap ... where earnings stop at age 62 but social security income doesn't begin until age 66.5 though? [In this post above] I understand that some people may choose to retire at age 62, draw on other resources, and then start taking SS at age 66.5, but that doesn't strike me as what the baseline assumption should necessarily be?
The recipient in Case A of the SS example was born in 1957 and therefore has a Normal Retirement Age (NRA) of 66.5. Case A page 2 calculates the Primary Insurance Amount (PIA) which is the benefit payable if one starts benefits at that age. I had to pick some age between 62 and 70 for benefits to start so I picked NRA.

According to this SSA web page for an NRA of 66.5, the benefit is 72.5% of the PIA if one starts at age 62 and 128% if one starts at age 70. Here is a comparison of the resulting rates of return for starting at 62, 66.5, and 70:

Code: Select all

``````Die   ---- Start SS @ ----
Age    62     66.5     70
---   ----    ----    ----
75   0.2%   (0.1%)  (0.8%)
80   1.3%    1.3%    1.2%
85   1.9%    2.1%    2.1%
90   2.4%    2.6%    2.7%
95   2.7%    3.0%    3.1%``````
And here are the year by year results for starting at 62 and 70. (The cash flows representing SS taxes paid are identical for all three cases. Only the benefits differ. See this post above for benefits and IRR for starting age of 66.5.)
Start SS at 62

Code: Select all

``````             Cash
Year  Age    Flow   IRR
----  ---   ------  ---
...
2017   60  (6,812)
2018   61  (6,882)
2019   62  16,354          = 72.5% X 12 X \$1,879.80 PIA
2020   63  16,354
2021   64  16,354
2022   65  16,354
2023   66  16,354
2024   67  16,354
2025   68  16,354
2026   69  16,354
2027   70  16,354
2028   71  16,354
2029   72  16,354
2030   73  16,354
2031   74  16,354  (0.1%)
2032   75  16,354   0.2%   <--
2033   76  16,354   0.5%
2034   77  16,354   0.7%
2035   78  16,354   0.9%
2036   79  16,354   1.1%
2037   80  16,354   1.3%   <--
2038   81  16,354   1.4%
2039   82  16,354   1.6%
2040   83  16,354   1.7%
2041   84  16,354   1.8%
2042   85  16,354   1.9%   <--
2043   86  16,354   2.0%
2044   87  16,354   2.1%
2045   88  16,354   2.2%
2046   89  16,354   2.3%
2047   90  16,354   2.4%   <--
2048   91  16,354   2.4%
2049   92  16,354   2.5%
2050   93  16,354   2.6%
2051   94  16,354   2.6%
2052   95  16,354   2.7%   <--
2053   96  16,354   2.7%
2054   97  16,354   2.8%
2055   98  16,354   2.8%
2056   99  16,354   2.8%
2057  100  16,354   2.9%``````
Start SS at 70

Code: Select all

``````             Cash
Year  Age    Flow   IRR
----  ---   ------  ---
...
2017   60  (6,812)
2018   61  (6,882)
2019   62
2020   63
2021   64
2022   65
2023   66
2024   67
2025   68
2026   69
2027   70  28,874          = 128% X 12 X \$1,879.80 PIA
2028   71  28,874
2029   72  28,874
2030   73  28,874
2031   74  28,874  (1.5%)
2032   75  28,874  (0.8%)  <--
2033   76  28,874  (0.3%)
2034   77  28,874   0.2%
2035   78  28,874   0.6%
2036   79  28,874   0.9%
2037   80  28,874   1.2%   <--
2038   81  28,874   1.4%
2039   82  28,874   1.6%
2040   83  28,874   1.8%
2041   84  28,874   2.0%
2042   85  28,874   2.1%   <--
2043   86  28,874   2.3%
2044   87  28,874   2.4%
2045   88  28,874   2.5%
2046   89  28,874   2.6%
2047   90  28,874   2.7%   <--
2048   91  28,874   2.8%
2049   92  28,874   2.9%
2050   93  28,874   3.0%
2051   94  28,874   3.1%
2052   95  28,874   3.1%   <--
2053   96  28,874   3.2%
2054   97  28,874   3.2%
2055   98  28,874   3.3%
2056   99  28,874   3.3%
2057  100  28,874   3.4%``````

Nate79
Posts: 5326
Joined: Thu Aug 11, 2016 6:24 pm
Location: Delaware

### Re: Estimating the rate of return on Social Security

How about running the rate of return if the \$50k was for a couple? But ignoring all the other insurance benefits greatly underestimates the value of SS.

Thesaints
Posts: 2921
Joined: Tue Jun 20, 2017 12:25 am

### Re: Estimating the rate of return on Social Security

You guys forget the insurance side of the SS program. That too is a benefit and it should be taken into account.

jeffyscott
Posts: 8442
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

### Re: Estimating the rate of return on Social Security

Thesaints wrote:
Sun May 05, 2019 5:47 pm
You guys forget the insurance side of the SS program. That too is a benefit and it should be taken into account.
Yes, at a minimum, the disability portion of the tax needs to be excluded when looking at the retirement benefits. Not doing so is overstating cost by about 17%.

In end the whole thing is an insurance program and nearly every cent of the "premiums" is or will be redistributed to the "policy holders".
Time is your friend; impulse is your enemy. - John C. Bogle

munemaker
Posts: 4147
Joined: Sat Jan 18, 2014 6:14 pm

### Re: Estimating the rate of return on Social Security

Given other factors being the same, the lower your income, the better your rate of return on your SS contributions.

#Cruncher
Posts: 2869
Joined: Fri May 14, 2010 2:33 am
Location: New York City
Contact:

### Re: Estimating the rate of return on Social Security

jeffyscott wrote:
Sun May 05, 2019 6:06 pm
... at a minimum, the disability portion of the tax needs to be excluded when looking at the retirement benefits. Not doing so is overstating cost by about 17%.
Workers do not have a choice to "contribute" to either the disability or to the old age parts of SS. They must pay into both. So I simply consider my results to be for the worker who does not become disabled and receives only the old age benefits.

munemaker wrote:
Sun May 05, 2019 6:27 pm
Given other factors being the same, the lower your income, the better your rate of return on your SS contributions.
Indeed! To illustrate this I repeated my calculation from this post of Case A in the SSA's Benefit Calculation Examples for Workers Retiring in 2019 using the 40 years of salary from Case B which are much higher. [1] The table below shows that doing this increased the PIA from \$1,879.80 to \$3,030.59 and proportionally increased the annual benefits when starting SS at age 62 and 70.

Code: Select all

``````                                          - Using Salary From -
Case A      Case B
---------   ---------
Top 35 years indexed salary               1,762,684   4,324,374
Average Indexed Monthly Earnings (AIME)       4,196      10,296
2019 90% to 32% bend point                      926         926
2019 32% to 15% bend point                    5,583       5,583
Primary Insurance Amount (PIA)             1,879.80    3,030.59
Benefit at age 62 (72.5% of PIA)             16,354      26,366
Benefit at age 70 (128% of PIA)              28,874      46,550``````
But this benefit increase is much smaller proportionally than the increase in salary on which it is based. [2] Since the SS taxes are proportional to salary, this makes the Internal Rate of Return lower for the worker with higher salaries. As the following table shows, the reduction ranges from 1.7% points (-1.5% vs +0.2%) for someone claiming at age 62 who dies at age 75 to 1.0% points (+2.1% vs +3.1%) for someone claiming at age 70 who dies at age 95..

Code: Select all

``````       --- Internal Rate of Return ---
------ Using Salary From ------
Die    -- Case A --       -- Csse B --
Age     62      70         62      70
---    ----    ----       ----    ----
75    0.2%   (0.8%)     (1.5%)  (2.3%)
80    1.3%    1.2%      (0.2%)  (0.1%)
85    1.9%    2.1%       0.6%    1.0%
90    2.4%    2.7%       1.1%    1.6%
95    2.7%    3.1%       1.5%    2.1%``````
Here is the detailed calculation for the worker with salaries from Case B who retires at age 70. Scroll down to see the Internal Rates of Return (IRR).

Code: Select all

``````             Cash          --- Salary ----            Tax
Year  Age    Flow   IRR    Case A   Case B   CPI-U    Rate``````

Code: Select all

``````1979   22  (8,473)         10,733   22,900   69.800  10.16%  [3]
1980   23  (8,351)         11,737   25,900   80.100  10.16%
1981   24  (9,128)         12,959   29,700   88.500  10.70%
1982   25  (9,413)         13,715   32,400   94.500  10.80%
1983   26 (10,011)         14,428   35,700   97.900  10.80%
1984   27 (10,676)         15,324   37,800  102.600  11.40%
1985   28 (10,785)         16,027   39,600  106.400  11.40%
1986   29 (11,187)         16,553   42,000  108.800  11.40%
1987   30 (11,323)         17,663   43,800  112.100  11.40%
1988   31 (11,901)         18,590   45,000  116.500  12.12%
1989   32 (12,092)         19,386   48,000  122.300  12.12%
1990   33 (12,564)         20,343   51,300  128.700  12.40%
1991   34 (12,468)         21,166   53,400  135.000  12.40%
1992   35 (12,559)         22,324   55,500  139.300  12.40%
1993   36 (12,644)         22,584   57,600  143.600  12.40%
1994   37 (12,977)         23,260   60,600  147.200  12.40%
1995   38 (12,742)         24,265   61,200  151.400  12.40%
1996   39 (12,693)         25,529   62,700  155.700  12.40%
1997   40 (12,884)         27,099   65,400  160.000  12.40%
1998   41 (13,292)         28,602   68,400  162.200  12.40%
1999   42 (13,869)         30,286   72,600  165.000  12.40%
2000   43 (14,038)         32,056   76,200  171.100  12.40%
2001   44 (14,383)         32,918   80,400  176.200  12.40%
2002   45 (14,967)         33,346   84,900  178.800  12.40%
2003   46 (14,888)         34,261   87,000  184.200  12.40%
2004   47 (14,785)         35,959   87,900  187.400  12.40%
2005   48 (14,676)         37,384   90,000  193.300  12.40%
2006   49 (14,861)         39,216   94,200  199.800  12.40%
2007   50 (14,966)         41,115   97,500  205.352  12.40%
2008   51 (15,057)         42,183  102,000  213.528  12.40%
2009   52 (15,827)         41,667  106,800  212.709  12.40%
2010   53 (15,469)         42,774  106,800  217.631  12.40%
2011   54 (15,065)         44,241  106,800  223.467  12.40%
2012   55 (15,129)         45,754  110,100  229.392  12.40%
2013   56 (15,397)         46,471  113,700  232.773  12.40%
2014   57 (15,608)         48,257  117,000  236.293  12.40%
2015   58 (15,819)         50,078  118,500  236.119  12.40%
2016   59 (15,686)         50,787  118,500  238.132  12.40%
2017   60 (16,446)         52,690  127,200  243.801  12.40%
2018   61 (16,218)         54,489  128,400  249.554  12.40%
2019   62                                   254.202
2020   63
2021   64
2022   65
2023   66
2024   67
2025   68
2026   69
2027   70  46,550
2028   71  46,550
2029   72  46,550
2030   73  46,550
2031   74  46,550  (3.0%)
2032   75  46,550  (2.3%) <--
2033   76  46,550  (1.7%)
2034   77  46,550  (1.2%)
2035   78  46,550  (0.8%)
2036   79  46,550  (0.4%)
2037   80  46,550  (0.1%) <--
2038   81  46,550   0.2%
2039   82  46,550   0.4%
2040   83  46,550   0.6%
2041   84  46,550   0.8%
2042   85  46,550   1.0%   <--
2043   86  46,550   1.1%
2044   87  46,550   1.3%
2045   88  46,550   1.4%
2046   89  46,550   1.5%
2047   90  46,550   1.6%  <--
2048   91  46,550   1.7%
2049   92  46,550   1.8%
2050   93  46,550   1.9%
2051   94  46,550   2.0%
2052   95  46,550   2.1% <--
2053   96  46,550   2.2%
2054   97  46,550   2.2%
2055   98  46,550   2.3%
2056   99  46,550   2.3%
2057  100  46,550   2.4%``````
1. The salaries in Case B are the same as the maximum amounts on which SS taxes are paid. Note that I did not use Case B in its entirety. I only used its salaries. To get the indexed earnings and PIA I used the same indexes and bend points as Case A. Thus I'm comparing two wage earners both born in 1957 who worked the same 40 years. Only the amounts of their salaries differ.
2. This is because of the way PIA is calculated. See page 2 of the two SSA examples.
3. Example cash flow calculation for age 22:
\$8,473 = 22900 * (254.202 / 69.8) * 10.16%

TeamArgo
Posts: 129
Joined: Fri Jul 22, 2016 8:18 am

### Re: Estimating the rate of return on Social Security

I am reminded of the joke about the physicist who was asked to look into low milk yields at a large dairy farm. A few hours later he was heard to say "I think I have a solution, if you assume a spherical cow in a vacuum..."

jeffyscott
Posts: 8442
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

### Re: Estimating the rate of return on Social Security

#Cruncher wrote:
Mon May 06, 2019 7:44 am
jeffyscott wrote:
Sun May 05, 2019 6:06 pm
... at a minimum, the disability portion of the tax needs to be excluded when looking at the retirement benefits. Not doing so is overstating cost by about 17%.
Workers do not have a choice to "contribute" to either the disability or to the old age parts of SS. They must pay into both. So I simply consider my results to be for the worker who does not become disabled and receives only the old age benefits.
Yes they must do both and they benefit from both, even if never disabled. If you want to look at the imaginary return on the total OASDI tax for some reason, it'd make more sense to include the cost of a private disability policy with equivalent benefits as part of your analysis.

I guess the actionable way to use this information would be to quit working early, so that you get a better "return". My spouse and I already implemented that option . While intentionally disabling yourself in order to improve your "return" would be fraud, I guess you could take up risky activities or get a risky job in order to do it .
Time is your friend; impulse is your enemy. - John C. Bogle

NotWhoYouThink
Posts: 2811
Joined: Fri Dec 26, 2014 4:19 pm

### Re: Estimating the rate of return on Social Security

No need to come up with some back of the envelope estimate on this, even someone with weak google skills can find dozens of well-researched articles on returns, taking into account low and high earners, people who die early or live past life expectancy ages, married or single, race, sex, and probably eye color.

As mentioned above, it is insurance, not investment. And it is there for people who have an earnings history but terrible money management skills - people who absolutely, positively would have spent the money instead of investing it for retirement if they had been allowed to keep it.

Stinky
Posts: 2854
Joined: Mon Jun 12, 2017 11:38 am
Location: Sweet Home Alabama

### Re: Estimating the rate of return on Social Security

TeamArgo wrote:
Mon May 06, 2019 8:05 am
I am reminded of the joke about the physicist who was asked to look into low milk yields at a large dairy farm. A few hours later he was heard to say "I think I have a solution, if you assume a spherical cow in a vacuum..."
One of the jokes in my company was when a senior executive asked an "unanswerable" question. The standard response to that was "What would you like the answer to be?"

In the extreme, a single worker with no dependent who works to age 69.5 and dies has a rate of return on his Social Security of 0.000000%. Paid into the system for 40-50 years, got \$0.00 in benefits.

It's an insurance program, not an investment program. Oftentimes the "rate of return" on insurance premiums is 0.000000% (if you never have a claim).
It's a GREAT day to be alive - Travis Tritt

Thesaints
Posts: 2921
Joined: Tue Jun 20, 2017 12:25 am

### Re: Estimating the rate of return on Social Security

#Cruncher wrote:
Mon May 06, 2019 7:44 am
Workers do not have a choice to "contribute" to either the disability or to the old age parts of SS. They must pay into both. So I simply consider my results to be for the worker who does not become disabled and receives only the old age benefits.
Then you results will be wrong.